Secure Detection of Android System by Jointly Selecting Features and Classifier Design

In order to improve the performance of android security detection, this paper puts forward a secure detection model of android applications by jointly selecting features and classifier design. Firstly, android security detection features are extracted, and the feature and support vector machine parameter are combined to a Individual, and then teaching learning optimization algorithm is used to simulate teacher's teaching process and students exchange process to find the best subset of features and support vector machine parameters, finally, the optimal android security detection model are established and the simulation experiment is carried. The experimental results show that, compared with other android security detection model, the proposed model has improved the accuracy rate of android security detection, and improved android security detection efficiency, so it can meet the real-time requirement of android security detection.